Biology Reference
In-Depth Information
promoters raises the question of whether coenrichment of these marks really
exists, or whether codetection simply reflects heterogeneity of the cell
populations in the embryo. In other words, do developmentally important
genes display multivalency in chromatin states or not? In ESCs (
Bernstein
et al., 2006; Mikkelsen et al., 2007
), as well as in somatic stem cells (
Cui
et al., 2009
) and differentiated cells (
Barski et al., 2007
), including the zebrafish
ZF4 fibroblast cell line (
Lindeman, Reiner, Mathavan, Alestr
¨
m, & Collas,
2010
), H3K4me3 and H3K27me3 appear to be coenriched on developmen-
tally regulated genes that are poised for later transcription, creating a “biva-
lent” domain (
Bernstein et al., 2006
). Claims of H3K4me3/H3K27me3
bivalency have been made based on sequential ChIP experiments, where
one modification is ChIPed and the other is in turn ChIPed from the eluate
of the first ChIP (
Bernstein et al., 2006; Lindeman, Reiner, et al., 2010
). Nev-
ertheless, many other studies claim bivalency on the basis of a mere
colocalization of two marks on a given locus in the absence of demonstration
of coenrichment. Intriguingly, H3K4me3/H3K27me3 coenrichment has
been reported exclusively in cultured cell lines and little evidence exists to
our knowledge to suggest that bivalency exists
in vivo
.
In embryos, the issue of bi- or multivalency of histone marks remains open
(
Andersen,
strup, et al., 2012; Vastenhouw & Schier, 2012
). In
Xenopus
gastrula stage embryos, H3K27me3-enriched promoters are devoid of
H3K4me3, arguing against bivalency (
Akkers et al., 2009
). Similarly, in
Drosophila
, there is essentially no evidence of H3K4me3/H3K27me3
bivalency, or if it occurs, its incidence is marginal (
Schuettengruber et al.,
2009
). A distinct finding arises from zebrafish post-MBT stage embryos, how-
ever, where rigorous sequential ChIP experiments suggest coenrichment
of developmentally regulated promoters by H3K27me3 and H3K4me3
(
Vastenhouw et al., 2010
). We have no such evidence of coenrichment in
zebrafish embryos in our hands, although this may be due to technical difficul-
ties associatedwith the low amount of material recovered after sequential ChIP
(Leif C. Lindeman and Philippe Collas, unpublished observations). The dis-
crepancy between findings in
Xenopus
,
Drosophila
, and zebrafish may be due
to species differences, developmental stages examined, and stringency and sen-
sitivity of the assays used to question bivalency. Conceptually, bi- or multi-
valency with coenrichment of permissive and repressive marks creating an
“idling” chromatin state constitutes an attractive mechanism whereby in
H3K27me3may keep lineage-specific genes in a temporarily repressed (halted)
state preventing unscheduled premature activation, while H3K4me3 provides
permissiveness enabling activation upon development and differentiation.
